The five ways Google Ads
quietly wastes your budget

First, a number to measure against

LocaliQ’s 2026 search advertising benchmarks put the average cost per lead across all industries at $66.69, on an average conversion rate of 8.18%. By category it swings hard: automotive repair around $29.96, dentists $72.97, home improvement $90.92, attorneys $131.63.

Use that as a sanity check, not a target. If your cost per lead is far above your category, something below is probably the reason. If it is far below, check what is being counted as a lead before you celebrate — that is usually the explanation.

1. Broad match without a negative list

Broad match is not the villain it was, and with good conversion data it can find volume you would never have thought to bid on. Without good conversion data it is an expensive random number generator.

The failure mode is predictable: a service business bids broadly on its core term and pays for searches about jobs, DIY instructions, wholesale suppliers, and the same service in a city three states away. The account looks busy. The search terms report tells a different story.

Open the search terms report for the last ninety days and sort by cost. Read the top fifty. In most accounts we audit, somewhere between a fifth and a third of spend has gone to terms the owner would never knowingly buy. Build the negative list from what you find, and keep adding to it monthly — it is not a one-time task.

2. Conversion actions that count the wrong thing

This is the most damaging one, because it corrupts everything downstream. If your conversion action fires on a page view, a click to the contact page, or every form submission including spam, then Google’s bidding is optimizing toward noise. It will faithfully find you more of whatever you told it to value.

We have seen accounts with a conversion rate above 40% that had produced almost no actual customers, because the conversion was “visited the contact page.” Smart Bidding then spent the budget chasing people who browse contact pages.

What should count: phone calls over a genuine duration threshold, form submissions that pass a spam filter, booked appointments. What should not: page views, clicks to a tel: link that never connected, newsletter signups counted at the same value as a sales enquiry. Fix this before touching bids, budgets or copy. Everything else is guesswork until it is right.

3. Paying for your own brand when nobody is competing for it

Brand bidding is genuinely necessary when competitors bid on your name, or when your organic listing is buried. It is pure waste when neither is true — you are paying for a click you were going to get free.

The test is simple. Search your business name in an incognito window, from a location your customers are actually in. If no competitor ad appears and your own listing is at the top, your brand campaign is mostly buying traffic you already owned. Pause it for two weeks and watch total conversions, not campaign conversions. If total holds steady, you found your answer.

4. Performance Max drifting into places you did not intend

PMax will spend across Search, Display, YouTube, Discover, Gmail and Maps, and it decides the split. Left alone it often drifts toward the cheapest inventory, which produces impressive impression counts and disappointing phone calls.

Three controls worth using: brand exclusions so PMax is not absorbing your brand traffic and claiming credit for it; account-level placement exclusions for the categories of app and content you do not want to appear next to; and asset groups built around real service themes rather than one group holding everything.

Then judge it on the same conversion actions as everything else. A campaign that cannot be compared on cost per qualified lead cannot be managed.

5. Landing pages that do not match the ad

The cheapest improvement in most accounts is not in the account at all. An ad promising emergency same-day service that lands on a homepage with nine services, three sliders and a phone number in the footer will convert at a fraction of the rate of a page about that one service, with the phone number in the first screen.

Match is three things: the headline echoes the ad’s promise, the primary action is visible without scrolling, and the page loads fast enough that people on phones do not leave first. That last one is its own subject — we go into what actually makes a site slow separately.

The order to fix them in

Conversion tracking first, always — every other decision depends on it being honest. Then negatives, because that is where the fastest money is. Then landing pages, which lift every campaign at once. Brand and PMax after that, because they need clean data to judge.

Working through this on your own account is a couple of afternoons. If you would rather we did it, it is the first thing we do on any paid media engagement — and it is the reason accounts in places like Garland and McKinney often get cheaper before they get bigger.